Li H, Yang P, Schwartz A G
Section of Biostatistics, Mayo Clinic, Rochester, Minnesota 55905, USA.
Biometrics. 1998 Sep;54(3):1030-9.
Age of onset data from case-control family studies are frequently used to study the relationship between disease and environmental or genetic factors. Dependence of age of onset within family and case-control sampling designs must be taken into account in analyzing such data. In this paper, we propose a parametric likelihood approach to study the relationship between disease risk and covariates, allowing for any correlations in age of onset of disease between family members and to study the familial association in age of onset of disease due to genetic factors, after adjusting for other shared covariates within the family. The method provides a way to combine the information relating disease incidence to risk factors in relatives with the information contained in the case/control contrasts in order to obtain more precise estimates of the effects of the putative risk factors. A data set from a case-control family study of lung cancer is used to illustrate the method. The analysis indicated that history of smoking, passive smoking, and history of chronic obstructive pulmonary disease are positively associated with lung cancer risk, and that, after adjustment for these risk factors, there is little evidence of familial aggregation in lung cancer risk.
病例对照家系研究中的发病年龄数据常被用于研究疾病与环境或遗传因素之间的关系。在分析此类数据时，必须考虑家系内发病年龄的依赖性以及病例对照抽样设计。在本文中，我们提出一种参数似然方法来研究疾病风险与协变量之间的关系，该方法允许家庭成员之间疾病发病年龄存在任何相关性，并在调整家庭内其他共享协变量后，研究由于遗传因素导致的疾病发病年龄的家族关联。该方法提供了一种将亲属中疾病发病率与风险因素相关的信息与病例/对照对比中包含的信息相结合的方式，以便更精确地估计假定风险因素的影响。来自肺癌病例对照家系研究的一个数据集用于说明该方法。分析表明，吸烟史、被动吸烟史和慢性阻塞性肺疾病史与肺癌风险呈正相关，并且在调整这些风险因素后，几乎没有证据表明肺癌风险存在家族聚集性。
